large diameter soft ground bored tunnel review

Cascadia Center

Large Diameter Soft Ground Bored Tunnel Review

Review of current industry soft ground bored tunnel practice

Document ref REP/208085/S001

Cascadia Center

Large Diameter Soft Ground Bored Tunnel Review

Review of current industry soft ground bored tunnel practice November 2008

This report takes into account the

particular instructions and requirements of our client. It is not intended for and should not be relied upon by any third party and no responsibility is undertaken to any third party

Arup North America Ltd 403 Columbia Street, Suite 220, Seattle, WA 98104 Tel +1 206 749 9674 Fax +1 206 749 0665 www.arup.com Job number 208085-00

Cascadia Center Large Diameter Soft Ground Bored Tunnel ReviewReview of current industry soft ground bored tunnel practice

C:\RICHARD\TUNNEL STUDY\REPORT_ISSUE.DOC REP/208085/S001

Arup North America LtdIssue 20 November 2008

Contents

Page 1 Introduction 1 2 Large Diameter Tunnels 2

2.1 Recently completed tunnels 2 2.2 The Future 3

3 TBM technology 5 4 Versatility of tunnel configurations 7 5 Whole life costing 8

5.1 ‘Like-for-like’ comparison 8 5.2 Project construction costs 8 5.3 Whole-life cost estimate 10 5.4 Broader benefits 11

6 Conclusions 12 7 References 13

Appendices Appendix A Large bored tunnel survey Appendix B Tunnel cross-sections Appendix C Extracts of referenced reports


C:\RICHARD\TUNNEL STUDY\REPORT_ISSUE.DOC REP/208085/S001 Page 1 Arup North America Ltd

Issue 20 November 2008

1 Introduction The Alaskan Way Stakeholder Advisory Committee is considering a number of options for the replacement of this important transportation corridor through Seattle.

Arup were commissioned by the Cascadia Center at the Discovery Institute to report on recent developments in the implementation of large diameter highway tunnel projects around the world.

This document reports on the technological advances that have made it common place to use large diameter bored tunnels for urban highway projects in difficult ground conditions, and reports on how increasing tunnel diameters make these subsurface corridors more versatile in terms of their function.

A discussion on costing highlights the importance of a whole life approach for such projects and how when considered in this way tunnels are typically more economic than other highway alternatives. Reported analyses publically available project construction cost data for tunnel projects around the world and in the United States also indicates that tunnel alternatives are competitive when compared with above ground options that provide similar capacity.

The document also highlights the wider economic, community and environmental benefits of tunnels that should be included in the assessment of replacement options.




2 Large Diameter Tunnels

2.1 Recently completed tunnels

The successful completion of two highway tunnels beneath the Yangtze River in Shanghai China in September this year represents another milestone in the development of large diameter bored tunnel projects around the world. The tunnels, the largest in the world to date at 51 ft diameter, will carry three lanes of vehicular traffic and a transit line in each direction between Pudong in Shanghai and Changxing Island (Figure 1), a distance of 4.7 miles. They were constructed in difficult ground conditions comprising sands and clays under high groundwater pressures some 200 ft below ground level. The tunnels were completed in 20 months and the project is expected to be open in 2010, one year ahead of schedule.

Figure 1 - Shanghai River Crossing (2 x 51ft tunnels)

This success with TBM’s of over 50 ft diameter comes on the back of the completion of the M30 tunnels in Madrid in 2008. These tunnels, at a diameter at 50 ft diameter, form part of a ring road around Madrid carrying three traffic lanes in each direction and were completed 4 months ahead of schedule. This subsurface project allowed the removal of highways from the banks of the Manzanares River and their development as a public amenity.

The 1.4 km Serebryany Bor tunnel (46 ft) in Moscow was also opened on December 27, 2007 becoming the first tunnel in the world to combine road traffic and metro transit.

These achievements reflect the improvements that have been made in tunnel boring technology over the last few decades and the increases in diameter that have recently been achieved. Table 1 indicates a selection of tunnels that are under construction or have been recently completed. This reflects the fact that such technology is widely used across the




world. It is also becoming more commonly used or proposed in the US for highway traffic such as the Port of Miami tunnels and the proposed I-710 project in Los Angeles.

Completed large diameter highway tunnels

Name Length Dia. Bores

Reported cost per mile of tunnel ($) Soils Function

Shanghai River Crossing, China 4.6 mi 50.6 ft twin $27m sand, clay, rubble Road

Madrid M-30 - north tunnel of the south bypass, Spain 3.65 mi 50 ft twin $131m

marly clays of the Madrid Tertiary penuela and gypsum Road

Serebryany Bor Tunnel 1.5 mi 46.6 ft twin no data no data Road/Metro

Lefortovo, Moscow 1.3 mi 46.6 ft twin $439m

fine to coarse sand, clay, limestone (medium strength, partially very fissured) Road

4th Tube of the Elbe Tunnel, Germany 1.6 mi 46.5 ft single $303m

sand and mud, rock and pebbles, marly till and mica schist Road

SMART Tunnel, Kuala Lumpur, Malaysia 1.86 mi 43.3 ft single $85m no data

Water/ Road

Wesertunnel, Kleinensiel, Germany 1 mi 38.3 ft twin $180m

clay, sand, turf, till, silt Road

Westerschelde, Terneuzen, Netherlands 4.1 mi 37 ft twin $60m

soft, permeable ground Road

A-86W East Tunnel, Paris France 6.2 mi 34 ft single $242m

limestone, sand, clay, marl, chalk Road

Table 1 - Completed large bore tunnels

See Appendix A for a list of completed, in construction and proposed large bore tunnels.

2.2 The Future

Looking to the future, large tunnels will become more commonly used and of larger diameter. On 27th March 2008, Moscow based ZOA Infrastruktura announced a deal to acquire a 62.3ft diameter tunnel boring machine. The ultimate project this is to be used for is undisclosed, but it will be the first bored tunnel capable of accommodating a four-lane highway.

Other projects now on the drawing boards around the world include bored tunnel alternatives up to 58 ft in diameter for the 4.5 mile long I-710 project in Los Angeles, three 50 ft TBM tunnels in Austria, a 50 m diameter 5th Elbe crossing tunnel, several large diameter transit tunnels for systems at Tysons Corner in Virginia, Sao Paulo and Barcelona. Figure 2 shows the increases in tunnel diameters that have been delivered over the last two decades alongside proposals which have been announced for the next 5 years.




TBM diameter development

ZOA Infrastruktura

Inntalquerung

Port of Miami

Groene Hart

Beacon Hill

Shanghai

Madrid M-30

4th Tube of the Elbe Tunnel

Lefortovo

New Southern Railw ay Project

WuhanWesterscheldeWesertunnel

SophiapoortunnelPannerdenschkanaal

SMART

Nanjing

A-86 West (E)

A-86 West (W)

Serebryany Bor

HERA

GrauholzZurich-Thalw ill

H3/4

Zaryzino Tunnel

H8

0

5

10

15

20

25

30

35

40

45

50

55

60

65

1985 1990 1995 2000 2005 2010

Tunn

el d

iam

eter

(ft)

Figure 2 - Increasing TBM diameters over the last 20 years




3 TBM technology Over the last two decades, the ability of TBMs to construct large diameter tunnels in weak ground and in significant groundwater pressures has increased significantly. Historically, tunnels in these ground conditions were constructed with the use of compressed air to provide support to the ground. Modern TBMs use a pressurized, or closed, face at the front of the machine to maintain stability in the ground while allowing workers to operate in safe conditions within the tunnel.

There are two main types of soft ground TBMs. An Earth Pressure Balance Machine (EPBMs) controls the rate at which the excavated ground is removed through the pressured face so that an appropriate level of support is provided to the ground around the tunnel, maintaining stability. A Slurry, or Mixshield, TBM uses a bentonite fluid to maintain the support and to carry the excavated ground in suspension. The choice between the two types of machine depends on the exact nature of the ground, with EPBMs more suited to clays and silt and Slurry machines to sandy ground or gravels. In complex projects, with varying ground conditions, a hybrid machine which has the ability to work in both modes can be used.

As a track record of successful tunneling with soft ground TBMs has been built up, and more sophisticated control systems have become available, the ability to excavate larger diameter tunnels in more challenging ground has increased. Some specific areas of improvement that have allowed soft ground tunneling to be carried out with less risk, lower impacts and with greater efficiency include:

• The motors used to rotate the cutting head at the front of the machine have been developed to become more powerful and can be controlled much more accurately, allowing larger TBMs that can operate in more complex ground.

• Modern soft ground TBMs can collect and process a vast amount of data in real time so that the operator has a good understanding of how the machine and the ground around it are performing, allowing a high level of control. This particularly allows very close control of ground movements above the tunnel, permitting the construction of tunnels in dense urban environments without damage to buildings or subsurface utilities.

• Increased scientific understanding of foams and additives that can be added into the excavated soil in the pressurized area at the front of the TBM, allowing the ground to provide a uniform support to the ground around the tunnel.

• Increased understanding of the abrasive effects of the ground on machines, allowing TBMs to be designed to suffer less wear and increase in reliability.

• Improved design of the cutting tools that are mounted on the front of the TBM, which provides better performance. Electronic sensors can also be incorporated in the cutting tools to detect when the tools are becoming worn down, which allows maintenance to be scheduled.

• On large diameter TBMs, the machines have been designed to allow the cutting tools to be replaced without sending workers into the pressurized zone in front of the machine, simplifying maintenance procedures.

• Improved design of seals in bearings and between the TBM and the tunnel lining have allowed tunnels to be constructed under higher ground water pressures.

• Ground penetrating radar has been used on some TBMs to help identify the location of boulders or other obstructions ahead of the tunnel, reducing the risk of unplanned stoppages.




These developments continue to evolve allowing ever more challenging conditions to be successfully tunneled.




4 Versatility of tunnel configurations With these increases in tunnel diameter, innovative configurations of corridors within the tunnels have been developed to optimize the usage of this underground space. These include double stacked traffic decks, separation of vehicle traffic from heavy goods vehicles, provision of transit corridors through highway tunnels and, in the case of the SMART tunnel in Malaysia, the use of part, and under some conditions all of the tunnel as a storm-water overflow tunnel. Figure 3 shows cross sections through three built tunnels which show this versatility. Appendix B contains figures from other tunnels referenced in this report.

Figure 3 – Road tunnel configurations (left Serebryany Bor, middle M-30, right A-86)




5 Whole life costing 5.1 ‘Like-for-like’ comparison

Cost estimating for large infrastructure projects is a complex process and is impacted by many variables. Cost comparisons of project alternatives in the US are typically carried out on the basis of construction cost only, however it is becoming more widely recognized that the longer term costs to an owner are also heavily influenced by the maintenance and operational costs and the eventual replacement costs or residual asset value. This type of analysis is common place in the manufacturing industry and is the approach taken by private investors when considering the construction or purchase of an infrastructure facility such as a highway.

The paper ‘Compiling real costs for true infrastructure construction comparison’ prepared for Tunnels and Tunneling magazine in December 2005 (Appendix C) reports that while the project cost for a bored tunnel may be slightly more than that of a viaduct and substantially more than a surface street option, when whole life costs are considered the tunnel becomes the most cost effective solution.

5.2 Project construction costs

Reported project construction costs for a number of tunnel projects are provided in Table 2. These figures have been compiled from a literature survey of completed and proposed projects. Data on major infrastructure projects is readily available. However, the basis of the estimates is often unclear. Tunnel project costs fall into the following categories:

• Tunnel contract cost (including cost of TBM and TBM operation)

• Project cost (including tunnel contract cost, support infrastructure, access roads, highway surfacing, etc.)

• PPP cost (includes project cost, operating and maintenance cost, cost of financing)

It is often difficult to arrive at a realistic cost estimate without some engineering defined criteria on which to base the costs. Baseline cost data from similar projects around the globe are useful but not always illustrative of the eventual costs of a project that do not contain the same combination of features that are required for the construction of a specific project.

With the above caveats in mind, Table 2 and Figure 4 summarize a literature survey of global large bore tunnel costs. As some projects comprise a single bore solution and others a twin bore solution the data is presented in dollars per tunnel mile. The calculation for a twin bored tunnel would be double the presented price per alignment mile. The highlighted projects form a cluster of projects which suggest a typical cost range per mile of a twin bore project of approximately $200M to $700M. It can be seen that projects at the lower end of the cost scale are often in Asian countries where labor and material costs are significantly lower. Higher costs per mile are reported in western European countries and in the US where the Port of Miami tunnel is approximately $1,000M for a 0.7 mile twin bore project. The shorter length of this project may have an upward impact on the cost per mile as economies of scale are smaller.

This data indicates that construction costs for tunnel projects larger than a 2 mile Alaskan Way by-pass tunnel are somewhat less than previously published estimates for the Alaskan Way Viaduct replacement.

In comparing these costs with estimates produced for the Alaskan Way project it should also be noted that project costs developed by WSDOT use a statistical approach to construction




costing (See ‘Probable cost estimating and risk management’ presented in Appendix C) which recognizes the various risks that may impact the project cost and applies a differing amount of contingency dependent on the certainty of the cost estimate. During project development, costs reported for a high level of certainty of non exceedance may be higher than those determined by more traditional methods.

Survey of bored tunnel reported costs (per mile of bored tunnel)

WesertunnelDublin Port Tunnel

Wuhan

Nanjing

M-30

I-710 (C3)

I-710 (A3)

Wuhan

A86W

Pannerdenschkanaal

Westerschelde

SMARTLefortovo

Shanghai River Crossing

Airport Link Brisbane

Beacon Hill Tunnel

Groene Hart Tunnel

Port of Miami Tunnel

15$0 $100 $200 $300 $400 $500 $600 $700 $800

Millions

Reported project cost (per mile of bored tunnel)

TBM

dia

met

er (f

t)

Figure 4 - Survey of tunnel costs




Tunnel Year

completed Diameter

(ft) Bores

Alignment length (miles)

Total length of tunnels

(miles)

Reported cost ($ million)

Cost per mile of tunnel

(million $/mile) Port of Miami Tunnel proposed 36 twin 0.7 1.5 1,000 $677

Lefortovo 2005 47 single 1.4 1.4 600 $439 Airport Link Brisbane 2012 41 twin 3.3 6.5 2,206 $338 Groene Hart Tunnel 2006 48 single 1.4 1.4 450 $332 4th Tube of the Elbe 2002 47 single 2.6 2.6 775 $303

I-710 (A3) proposed 501 triple 4.1 12.4 3,585 $290 I-710 (C3) proposed 421 triple 4.0 12.0 3,195 $266

A86W 2010 37.91 single 10.9 10.9 2,641 $242 Wesertunnel 2001 38 twin 1.0 2.0 358 $180

Beacon Hill Tunnel 2009 21 twin 0.8 1.6 280 $172 M-30 2008 50 twin 2.2 4.3 570 $131

Dublin Port Tunnel 2006 38 twin 2.8 5.6 530 $94 Pannerdenschkanaal 2003 32 twin 1.0 2.0 173 $86

SMART 2007 43 single 6.0 6.0 515 $85 Wuhan 2008 37 twin 1.7 3.4 288 $85 Nanjing 2013 49 twin 1.9 3.7 245 $66

Westerschelde 2002 37 twin 4.1 8.2 490 $60 Shanghai River

Crossing 2008 51 twin 4.6 9.3 245 $27 1 This scheme contains multiple tunnel diameters. This number presented is the average tunnel diameter.

Table 2 - Survey of tunnel costs

5.3 Whole-life cost estimate

As indicated by the Tunnels and Tunneling paper (Appendix C) maintenance and operational costs for a road tunnel are likely to be similar to that of a surface street option and significantly lower than that of an elevated highway. The maintenance of way costs will be similar, however maintenance and operation of ventilation systems can be an additional cost of longer tunnels while structural repair cost for viaducts typically increases that of an above ground structure.

The largest whole life benefit of a tunnel is its longer design life. The life span of tunnels can be up to 150 years against 50 years for an elevated structure. This is evidenced in Seattle by the BNSF tunnels which are over 100 years old against the existing Alaskan Way viaduct and SR-520 which are reaching the end of their life span after approximately 50 years.

It should also be noted that, given the potential for seismic events in the Pacific North West, tunnels perform better in earthquakes as evidenced by the damage caused to the Bay Bridge and the Embarcadero highway in San Francisco against the performance of the BART transit tunnels below the bay which remained largely unaffected by the Loma Prieta earthquake in 1989.

The importance of whole life costing is demonstrated in Table 3 below which illustrates that the “conventional” cost of an infrastructure project could amount to only 26% of the whole life cost. A full whole-life cost of any infrastructure replacement option is therefore the most appropriate approach to generate a fair and comparable assessment.




Table 3 - Whole life cost

5.4 Broader benefits

Major infrastructure projects typically have a wider benefit to the community and as such these benefits should be fully considered in the assessment of options. The benefits of a tunnel option, when compared with above ground options include increased property values, reduced congestion and associated pollution, and reduced disruption to businesses, pedestrians and traffic during construction. Land freed up by placing the facility below ground presents opportunity for improved public amenity and development opportunities particularly at the portal locations. These benefits can be assigned a monetary value which should be included in a detailed assessment.

The impacts of these issues in relation to comparing a tunnel with a surface street and a viaduct are discussed in the Tunnel and Tunneling article presented in Appendix C. This article makes two significant points:

a) tunnels have the lowest pro-rate cost taking into account the longer design life of tunnels.

b) when the cost of environmental pollution, loss of property taxes, social divide and maintenance cost is taken into account, the annual costs of a typical tunnel is half that of the at grade option and a third that of an elevated structure.




6 Conclusions Bored tunnel technology is continually developing and larger and more safely constructed tunnels are becoming more and more common. This additional size allows more versatile configurations of uses within the tunnel to allow stacking of traffic lanes and the combination of different uses such as highway and transit usage.

The total whole life cost for a project, rather than construction costs alone, better reflect the overall costs incurred by an owner for a transportation facility. Methods for assessing whole life costs are available and are commonly used in the private sector for assessing such projects.

Major transportation projects have a wider impact on the surrounding community. Tunnels typically provide greater benefits than other alternatives as a result of:

• Reduced disruption to businesses during construction

• Reduced utility relocations,

• Reduced street impacts to pedestrians and vehicles during construction

• Increased property values as result of the infrastructure being below ground

• Greater opportunity to provide public amenity

• Development opportunities where land is released by placing facility below ground particularly at portals

• Providing capacity below ground moves congestion and pollution from downtown streets.




7 References “Tunnel Boring Machine Development”, Martin Herrenknecht and Karin Bäppler, North American Tunneling Conference 2008

“State of the Art in TBM Tunneling”, Martin Herrenknecht and Karin Bäppler, North American Tunneling Conference 2006

“Current Issues Regarding Mechanized and Automated Tunneling”, G.Ishii, Tunnels and Underground Structures 2000

“Lifting the lid on Mixshield performance”, Werner Burger and Gerhard Wehymeyer (Herrenknecht AG), Tunnels and Tunneling International June 2008

“Compiling real costs for infrastructure construction comparison”, T&T North America, December 2005

“Probable Cost Estimating and Risk Management”, John Reilly

“Bold and Visionary Planning of Tunnels and Underground Space”, Harvey W. Parker

Appendix A Large bored tunnel survey


C:\RICHARD\TUNNEL STUDY\REPORT_ISSUE.DOC REP/208085/S001 Page A1 Arup North America Ltd


Name Length Diameter Bores Status Period Soils Function

I-710 Option A3 4.12 mi 58 ft

triple (1 x 58' and 2 x 46') proposed

shale, sandstone, siltstone, and conglomerate, with soft alluvial soils Road

I-710 Option A1 4.12 mi 57 ft twin proposed

shale, sandstone, siltstone, and conglomerate, with soft alluvial soils Road

Shanghai River Crossing 15.5 meter Soft Ground ( Bouygues Job) ( just finished) 4.6 mi 50.6 ft twin

tunnels completed 2008 sand, clay, rubble Road

Madrid M-30 - north tunnel (left carriageway) of the south bypass 2.2 mi 50 ft single completed

2004 - 2008

marly clays of the Madrid Tertiary penuela and gypsum Road

Nanjing Yangtze Crossing 1.8 mi 49 ft twin

contract awarded - 2013

oft river deposits of clay, silt and sand Road

Groene Hart Tunnel 5.3 mi 48 ft single completed 2000-2006

highly permeable sand below a very soft peaty clay layer Rail

Zaryzino Tunnel 1.5 mi 46.6 ft single proposed no data no data Road

Serebryany Bor Tunnel 1.5 mi 46.6 ft twin tunnel complete -2007 no data Road+/Metro

Lefortovo, Moscow 1.3 mi 46.6 ft twin complete 2000 - 2005

Fine to coarse sand, clay, limestone (medium strength, partially very fissured) Road

4th Tube of the Elbe Tunnel, Hamburg 1.6 mi 46.5 ft single complete

1997 - 2002

sand and mud, rock and pebbles, marly till and mica schist Road

SMART Tunnel, Kuala Lumpur, Malaysia 1.86 mi 43.3 ft single completed - 2007

Water/ Road

Tyson' Corner 3.38 mi 43.3 ft single proposed no data

residual soil, decomposed rock, and rock

Inntalquerung 3.6 mi 42.6 ft single in construction no data

Pebble stones, sand, coarse clay, brash, gravel Rail

Airport Link Brisbane 3.3 mi 41 ft twin tendered - 2012 no data Road

San Vito 0.6 mi 40.3 ft no data no data no data no data Rail

Rennsteigtunnel 5.6 mi 40.3 ft twin complete 1998 - 2003 no data Road

Metro Barcelona Line 9 5.3 mi 39.6 ft single completed no data

Granite, sand, clay, gravel, gravel with boulders Metro


C:\RICHARD\TUNNEL STUDY\REPORT_ISSUE.DOC REP/208085/S001 Page A2 Arup North America Ltd


Name Length Diameter Bores Status Period Soils Function

Hsuehshan (also known as Pinglin), Taiwan 8.0 mi 38.5 ft twin complete

1993 - 2006

strong, hard, abrasive and intensely fractured Szeleng Sandstone on the east and indurate sandstone and siltstone on the west Road

Wesertunnel, Kleinensiel, Germany 1 mi 38.3 ft twin complete

2000 - 2001

clay, sand, turf, till, silt Road

Wuhan, China 1.7 mi 37.3 ft twin in construction 2004 - 2008

sand stratum, gravel, and shale rock. Road

Westerschelde, Terneuzen, Netherlands 4.1 mi 37 ft twin complete

1998 - 2002

soft, permeable ground

Katzenberg 5.6 mi 36.5 ft twin complete 2003 - 2008

Mudstone, marl, limestone and sandstone, approx. 800m of Oxford Coral Limestone Rail

Port of Miami Tunnel 3900 ft 36 ft twin proposed no data no data Road

A-86W West, Paris 4.7 mi 35.8 ft single in construction 2007 - 2010


Finnetunnel 4.23 mi 35.5 ft twin in construction 2008 - 2009

loose rock formations in the first section of around 1,500-m-long and unsupported rock on the remaining line (sandstone, claystone) Rail

New Southern Railway Project 3 mi 35 ft single complete

1996 - 2000 Clay, sandstone Metro

Metrotren Gijon 2.4 mi 34.6 ft single completed - 2006 Clay, dolomit, limestone Metro

A-86W East 6.2 mi 34 ft single complete 2003 - 2008


Sophiaspoortunnel, Rotterdam, Netherlands 2.6 mi 32 ft twin complete

2001 - 2002 clay, sand, till Rail

Pannerdenschkanaal Tunnel, Netherlands 1 mi 32 ft twin

tunnel structure complete

2000 - 2003 sand, clay Rail

Sao Paulo metro , Linea 4 4.7 mi 31 ft single in construction - 2008 no data Metro

Leipzig , Germany 0.9 mi 29.5 ft no data in construction

2006 - 2008

no data low subsidence requirement below existing buildings Rail

Beacon Hill Tunnel 4,300 ft 21 ft twin in construction -2009 glacial soils Metro

Appendix B Tunnel cross-sections


C:\RICHARD\TUNNEL STUDY\REPORT_ISSUE.DOC REP/208085/S001 Page B1 Arup North America Ltd


Serebryany Bor Tunnel Moscow, Russia 1.5 miles 46.6 ft diameter

SMART Tunnel Kuala Lumpar, Malaysia 1.86 miles 43.3ft diameter

M-30 Madrid, Spain 3.65 miles 50 ft diameter




Nanjing Yangtze Crossing China 1.8 miles 49 ft diameter

A-86W East Paris, France 6.2 miles 34 ft diameter (note double deck configuration is limited to cars only)

A-86W West Paris, France 4.7 mi 35.8 ft diameter




Wuhan China 1.7 miles 37.3 ft diameter

Appendix C Extracts of referenced reports

489

Bold and Visionary Planning of Tunnels and Underground Space

Harvey W. ParkerHarvey Parker & Associates, Inc., Bellevue, Washington

INTRODUCTION

Our industry has developed to the skilled, experi-enced and productive industry that we now enjoy as aresult of the visions of great engineers throughout his-tory, but, in particular for their visions in the last 1centuries. Unfortunately for our forefathers, the tech-nology lagged so far behind that these exceptionalvisions could not be realized for decades, sometimesfor a century. All that has changed in recent time andmany bold and visionary projects can now be con-structed and operated because of dramatic advancesin technology. Before, say, 1950, tunnel technologydeveloped very slowly. However, technology is nowdeveloping at such a fast rate that planners and deci-sion makers are challenged to stay ahead. Moreover,new technology will develop during the lifetime ofany project that can dramatically improve under-ground space projects. One of the best examples ofthis technological advancement is that of closed-faceTBMs which now make tunneling feasible in previ-ously impossible ground conditions.

Since it usually takes a decade or more to plan,overcome environmental and political issues, con-struct, and put into service major projects, plannersand designers are challenged by the changes in tech-nology that occurs during the long time span ofplanning and construction. This is particularly trueof large infrastructure projects such as transporta-tion, water and wastewater systems and other majorunderground schemes. Also, since the serviceablelife of many of these tunnels is often over one cen-tury, planners must consider not only lessons learnedfrom the past but also what new concepts and inno-vations may develop over an extremely long time.These and other issues create a great challenge toplanners and illustrate the great importance of care-ful but creative vision and flexibility during planningof long-term tunnel and underground space projects.

PLANNING CHALLENGES POSED BY WORLD POPULATION

The world’s population is increasing and is re-group-ing at a staggering pace. In October, 1999, the worldpopulation passed the 6 Billion mark. However, amajor factor in world demographics that is extremelyimportant to planning of underground space is that, inthe future, most of the world’s population will live,

not in rural areas, but will live in urban cities. In 1950,only about 1⁄ 3 of the world’s population lived in urbanareas but that is changing at a rapid rate. By October,1999, about of the 6 Billion people lived in urbanareas. The trend continues to accelerate and cities arebecoming extremely large. So large, that the UnitedNations and other world bodies have given specialattention to those Megacities in which more than 10Million people live. In 2001, there were only 19 Meg-acities around the world. By 2015 it is estimated thatthere will be about 60 Megacities and most of thesenew Megacities will be in the Developing World. Thetrend will continue. By 2030, it is estimated that 4.9Billion people will live in cities which is 60% of theestimated 8.1 Billion world population.

Planners and decision makers must realize thatan enormous amount of infrastructure must be con-structed not just for the urban framework of large andvery large cities to be sustainable, but just for themjust so they can survive. In fact, there may not beenough trained and experienced planners and tunnel-ers to safely construct and operate such a large num-ber of underground facilities in such a short time.

Fortunately for the underground space indus-try, the underground is often the best location formuch of the infrastructure required in urban areas,especially if the environment and sustainable devel-opment are considered. It is essential that those of usin the tunnel and underground space industry be pro-active to inform the public, the media, city officials,and planners, at very early stages of a city’s growth,of the importance of the underground to sustainabledevelopment and to quality of life.

Greater use of underground space will be benefi-cial to sustainability of cites. This is particularly trueof essential infrastructure such as water, wastewater,other utilities and transportation. However, moreplanning and promotion of general undergroundspace such as bulk storage, manufacturing, living,office spaces will be important but also long tunnelswill be needed to maintain the efficiency of the trans-portation networks between these Megacities.

ENVIRONMENTAL BENEFITS AND THEIR IMPACT ON PLANNING

Tunnels and underground space play a dominantrole in our standard of living and in the preservation

490

of the environment. Moreover, they are increasinglyrecognized as being environmentally friendly andenergy efficient. But the advantages of the under-ground are taken for granted by almost everyone.Tunnels and underground space are “Out-of-Sight,Out-of-Mind.” At least within developed urban cit-ies, every time someone turns on a faucet or flushesa toilet, an “environmental tunnel” is put into use.Yet the average person almost never makes this con-nection; even members of our industry take cleanwater in and wastewater out for granted because ourtunnels have been doing such a good job for such along time with so little maintenance that they areinvisible to sight and to mind. As urban centers andmegacities develop, underground space will play anincreasing role in improving the quality of life.Underground specialists must play an important rolein the future of our cities by advising urban plannerson the advantages of underground space. In Decem-ber 2008, ITA conducted a joint UN-ITA workshopat the United Nations titled “Underground Space, anUnexpected Solution to Promote Sustainable Devel-opment.” There was considerable interest shown bythe UN delegates and audience and subsequentcooperative events are being planned (ITA 2007).Some of the numerous environmental and societalbenefits that are inherently associated with tunnelsand underground space are given below (Parker2004, ITA 2007).

• Tunnels play a vital role by conveying cleanwater to urban areas and by conveying wastewater out. Most major urban areas depend ontunnels for these services, which function witha minimum of maintenance.

• The usable space off a parcel of land can, insome cases, be almost doubled by adding trans-port or other communication lines, floor spaceor bulk storage below the ground surface.

• Life-cycle cost and benefit methodologies areexpected to reveal the underground alternativesto be much more competitive with surfacefacilities.

• It has been demonstrated by several recentevents that tunnels behave very well in earth-quakes. Underground space provides strongprotection from natural and man-made hazardssuch as Earthquakes, Storms, and Violence so ifyou want to protect or preserve an asset for thefuture, put it underground.

• Underground space inherently conservesenergy and thus promotes sustainable devel-opment and is an ally in the fight against Glo-bal Warming. Only a little energy is needed tobring underground space to a comfortabletemperature both in winter and summer.

Transportation Tunnels that reduce travel timeover rough mountainous terrain also provideenormous fuel savings and reduction of emis-sions for decades.– Underground space provides significant

environmental benefits by efficiently put-ting essential services underground leav-ing the surface to be used for more noblepurposes.

– Underground Space is Good for Sustain-ability and is an Ally in the fight againstGlobal Warming.

These advantages are very compelling espe-cially our industry’s contribution to the fight againstGlobal Warming. It will be up to the leaders of ourindustry to inform planners and decision makers ofthese advantages and benefits and, as they becomebetter known, there will be an enormous demand forconstruction and utilization of tunnels and under-ground space.

CHALLENGES TO PLANNERS AND DECISION MAKERS

What are some of other challenges facing Plannersand Decision Makers? They are not only numerousbut they also change in significance with time as theprocess of planning develops and becomes moremature, detailed, and comprehensive. Selected plan-ning challenges are outlined below:

• Population and Demographics• Economic and Social Values• Environmental Values• Tunnel Technology likely to be available• Non-Tunnelling Technology Developments• Cost and Funding issues

These, and other issues, need to be addressedearly in the planning of any underground project.Moreover, many of these issues are non-technical.Thus the planner is faced with multi-disciplinarydecisions from the outset and throughout the rest ofthe project. For instance, in the decade that it maytake to bring a project to construction, many of thebasic decisions may change. Demographics maychange and, in fact, may change just because thetunnel project is built and put in to service.

Changing environmental and social values arealso likely to result during the planning and construc-tion period of any project. With the rapid awareness ofthe environment, sustainable development and globalwarming, it is likely that environmental requirementsfor underground facilities will become stricter. On theother hand, the environmental benefits of tunnels may

491

just make the project more acceptable to the environ-mental community.

Of great importance to the technical tunnelcommunity will be the future changes of tunnelingtechnology and in the non-tunnelling technology.These changes are likely to be very important to theplanning and construction of underground facilities.The real challenge for planners will be to identifysuch technological improvements in time to applythem on the project.

Financial cost and funding issues are alwaysthe governing issues. These also change with timebut, unfortunately, the general trend is more tight-ened budgets.

DEVELOPMENT OF UNDERGROUND TECHNOLOGY IN THE PAST

General Comments on Development of Underground Technology

Clearly, major mining and civil works schemes dur-ing ancient times and throughout the Middle Ageswere bold and visionary feats in their own time. Thedevelopment of tunnelling technology was very slowfrom 1850 to almost 1950. Before 1950, technologi-cal developments were slow to materialize and tun-nel planners likely realized that there was not muchchance of significant innovations that might impacttheir project. Now (2008), just the opposite is true;innovations are now expected and rapid develop-ment in technology and innovations must be takeninto account by planners imposing a whole new setof visionary requirements for tunnel planners.

The author was fortunate to have been involvedin pioneering work on innovative tunnel support sys-tems in the late 1960s at the University of Illinois inChampaign-Urbana for a project for the U.S. Fed-eral Railroad Administration (Parker and Semple1972). The impetus for this work was the ambitiousconcept of a tunnel or series of rail tunnels fromWashington DC to Boston, a distance of some750 km. The concept was bold and visionary butwas never constructed for many reasons includingcost. As we undertook this project on innovations,the project team found that very few ideas were trulynew; almost everything had been thought of andtried before. Planners should understand that manynew ideas are not actually new ideas but are olderideas being revisited.

Since technological changes were slow todevelop, tunnel planners probably recognized thatthere was not much chance of significant innova-tions that might impact their project. This made theplanner’s job easier but the situation is reversed nowgiving tunnel planners great challenges.

Pre-1950 Development of Underground Technology

Our forefathers were great engineers. Even prior to1900, practicing tunnel planners, designers, and con-tractors were visionary in their development of someextraordinary concepts for tunnel construction andoperation. They not only had great visions but theyactually tried some of these great ideas. For no faultof their own, more frequently than not, they wereunsuccessful, mainly because technology had notprogressed far enough to support their ideas andvisions. Often those same ideas were successfullyput into practice decades or sometimes a centurylater (Parker, 1999). For instance, the Channel Tun-nel took over a Century to realize the dream, not onlyfor political reasons but also for technical reasons.

The first mechanized tunnel machine was triedunsuccessfully in the late 1840s on the 12 km longFrejus Tunnel (also known as the Mt. Cenis Tunnel)from Italy to France (Stack, 1982). The secondmechanized tunnel machine, also unsuccessful, wasbuilt for the 8-km-long Hoosac Tunnel in Massachu-setts, USA in the early 1850s. Over the next centurymany attempts to develop a mechanized tunnellingmachine were unsuccessful. However, there was atime-span of over a century before TBMs began tobecome practical.

Gunite, the forerunner to shotcrete was firstdeveloped in the USA around the turn of the centurywhen Carl Akeley invented the Cement Gun to applymortar over skeletal frameworks of prehistoric ani-mals for Chicago’s Field Museum. Although tried inan experimental mine in Pittsburgh in 1914, itwasn’t until 1952 on the Swiss Maggia Hydroelec-tric project before tunnels were solely supported byshotcrete (Parker, 2001). Again, there was a centuryspan between vision and practical application.

Development of Technology: 1950 to Present

Technology is advancing now (2008) at an everincreasing rate. Around the 1950s and beyond, tech-nology started to catch up with our vision whichresulted in many of our forefather’s ideas becoming areality. James S. Robins built an 8-m-diameter rotaryTBM which successfully excavated Pierre Shale atOahe Dam in the USA. He then pioneered the use ofsolely using disc cutters on a TBM cutterhead on asmall diameter tunnel in Toronto. Unfortunately,because of technological limitations, similar successwas not possible in hard rock for a very long time.

About a century after Greathead’s 1874 patentof a shield with water pressure on the face, pressur-ized-face TBMs became feasible. In the 1960s, sev-eral developments took place all over the world thatmade these machines more feasible technically.These took place in Japan, Germany, England, Mex-ico, USA, Canada and elsewhere. Earth Pressure

492

Balance and Slurry machines now are commonplaceon routine projects. Now, there are various conceptsfor handling mixed-face conditions to makemachines more feasible for rock tunnels that willalso pass through fault zones with differing groundconditions.

Steel fiber reinforced shotcrete (SFRS) wasdeveloped in the USA around 1970. The authorworked on the practical development of SFRS at theUniversity of Illinois in the early 1970s (Parker1976, Parker 2001). It only took a decade beforeSFRS became accepted and now (2008) SFRS isnow routinely used in tunnel construction.

Accordingly, now, because of such greatadvancements in technology, the ideas and vision ofthe planners, designers, manufacturers and tunnelersin our industry are being realized in the field veryquickly after the idea is conceived. This trendseemed to have started in Japan where some amaz-ingly creative machine configurations and conceptshave been built and used in practice but the trend isnow beginning to spread worldwide. The message tothe planners and designers is that our industry isvery creative and can overcome any challenge. Thus,planners should be bold and daring in the planningof projects.

Previous Planners Lacked the Tools

In the 1950s and 1960s, planners hardly had any ofthe tools that we now enjoy (2008). It is interestingto reflect on the tunnels that could have been built ifour predecessors knew that new and innovative toolswould exist when their tunnel was finally con-structed. In order to reflect on just how much ourindustry has changed in the last half-century, it ishelpful to reflect on the progress our industry hasmade since 1950. Equipment such as rock TBMs,Pressurized-face TBMs, conveyor belts, microtun-neling or directional drilling all have been developedin that time period since 1950. Methods of excava-tion and support such as the sequential excavationmethod (SEM/NATM), rock bolts and shotcrete,special ground improvement, single pass concretesegments and waterproof membranes also have beendeveloped since 1950.

Each of these (and the many others not listed)make tunnels and underground construction todaymore technically and financially feasible. Had plan-ners known that such innovations would becomeavailable, they could have been more bold andvisionary and many more tunnel projects wouldhave been proposed and built in increasingly moredifficult conditions. Although this is Monday morn-ing quarterbacking, it is proof of the strength andcreativity of our tunnelling industry. If a planner candream it up, our industry can get it done. Of coursethe difficulty is knowing what innovation is most

likely to become practical during the decade or so ofthe planning process. This is a task that UCA andITA should work on.

IDENTIFICATION OF OPPORTUNITIES AND RISK MANAGEMENT

Risk management has now become a buzz word inour industry and it is the subject of numerous con-ferences, books and papers. Tunnel owners and plan-ners now have begun to use systematic riskmanagement principles to identify all risks in a waythat directs the rest of the planning and constructionprocess to minimize those risks. This systematicprocedure must be done as early as possible in thestages of a project (pre-conceptual or idea stage).This systematic risk management work then is car-ried on and updated all the way through design andconstruction. The risks to be considered should bebroad and also include risks of cost, schedule, envi-ronment, public acceptance, adjacent owners andthird-party intervention, politics, etc. in addition tothe technical risks that always immediately come tomind.

Fortunately, the same concepts and tools can beused to identify value engineering ideas, as well asto identify broad ideas and opportunities including“thinking out of the box” (Parker and Reilly 2008).

PRINCIPLES OF LIFE-CYCLE COSTS AND BENEFITS

Tunnels often remain in service for over a century.Accordingly, decisions about whether a certaininfrastructure should be a tunnel, or not, should bemade on considerations of Life-Cycle Cost and Ben-efit and not Initial Capital Cost. This is a difficultconcept to implement but it is important for plannersand decision makers to avoid the pitfall of decisionsbased on initial capital cost. Using principles similarto those used in Risk Management, the likely cost ofa tunnel or underground facility and also its planningand construction schedule should be developed andreported as a range, not as a single number (Reilly,2006, Reilly and Parker 2007, Parker and Reilly2008).

Obviously, the life-cycle costs should includefuture operational and maintenance costs. However,the cost analyses should also include realistic allow-ances for equivalent financial benefits from environ-mental and social improvements associated withtunnels which can be substantial.

Thus, the initial cost is mitigated and offset bysignificant savings attributed to the environmentover the years of operation. It is imperative for plan-ners to consider the financial aspects of tunnels froma Life-Cycle Cost standpoint that also takes into

493

account the accumulative equivalent financial bene-fits from saving the environment. Planners and deci-sion makers must be courageous and convincing asthey present a clear message of the overall advan-tages of tunnels and underground space to themedia, public, politicians, and fellow decision mak-ers. UCA and ITA should take this challenge on as amajor goal.

BOLD AND VISIONARY TUNNEL CONCEPTS

Visionary Concepts Currently Being Implemented

There are many projects around the world that fallinto the category of bold and visionary. In Seattle,Washington, a 50-year-old viaduct along the water-front was damaged by an earthquake and needsreplacing. However, in addition, the old seawallalong the waterfront and just beneath the viaductalso needs replacing. Both structures may not sur-vive the next big earthquake. Planners and decisionmakers have designed a cut and cover roadwaywhose outboard wall will be designed to also func-tion as the new seawall. This new structure wouldreplace both the viaduct and the seawall with onestructure with an overall reduction in cost, schedule,and especially in reduced disruption to the public.Moreover, the cost would be shared by both roadand waterfront authorities. Of course, the initial costof any such structure is considerable and, at thepresent time, other schemes are being considered.Unfortunately, the comparison of schemes has been,and will likely continue to be, on the basis of capitalcost, not life cycle cost and benefits which also takeinto account the environmental and societal benefitsof the underground option.

Another innovative out-of-the-box concept isthe SMART tunnel project in Kuala Lumpur. It is adouble-deck tunnel is specially configured to handleboth auto traffic and floodwater. During low andmedium flows, water flows beneath the lower deckwhile cars are still travelling through the tunnel.However, when a very big flood occurs, traffic isremoved and the flood waters pass through the entiretunnel including the roadway. This way, the publicgets two end uses for the tunnel for a price and con-struction disruption that is less than that of two sepa-rate tunnels. Moreover, the cost of the tunnel isshared by two groups making each easier to afford.The concept of using tunnels to store wastewaterduring a storm, such as the Chicago TARP projectand other CSO storage projects, is another creativeuse of tunnels. Similar wastewater storage projectsare being constructed worldwide.

A pioneering concept is the A86 road tunnelproject in Paris. This project will not permit trucksand is strictly for cars less than 2-m-high. With thisrestriction, the owner/concessionaire is able to fit4 lanes of traffic plus 2 breakdown lanes (in a dou-ble-deck configuration) in a 11.6 m-outside-diame-ter tunnel. In the USA and elsewhere in the world,geometrical restrictions would only allow 2 lanes oftraffic. There is even a possibility for future expan-sion to have 3 lanes on each deck for a total of6 lanes in a tunnel. Such a configuration at leastgives the potential of the cost of such tunnels per kmper lane to be on the order of 1⁄ 2 to 1⁄ 3 of the cost fortraditional configurations. Moreover, such a tunnelcan be constructed using more readily availablestandard size TBMs in a shorter construction timeand with less disruption to the public.

A welcome breakthrough in the planning ofutilities has been achieved in Ashgabat, Turkmeni-stan where the planners and decision makers havefinally been able to implement a 30-km multi-useutility tunnel scheme. This 6 meter diameter tunnelacts to lower the groundwater table by acting as adrainage collection and transfer tunnel which col-lects and stores unwanted subsurface drainage waterand transfers it to discharge points in the desert. Thesame tunnel contains electrical power lines, tele-phone conduits and potable water pipes in differentcompartments (Wallis 2007). Our industry hastalked about such multi-use tunnels, so called utili-dors, for decades and generally they are rejected forlegal liability reasons. It is good to see such a projecton such a scale being constructed in Turkmenistan.

Bold, Visionary, and Daring Concepts

There are numerous very bold concepts that havebeen proposed. Many appear almost outrageous andmay never be built because of cost or other reasonsbut they have been proposed and talked about byserious planners and engineers.

A Submerged Floating Tunnel (SFT) has beenproposed several times in various parts of the worldbut never built. Conceptually, the SFT can be a lotshorter tunnel which gives much greater flexibilityin locating tunnel alignments and portals. SFTs are aperfect example of “thinking out of the box.” It is socreative that one can allow your imagination to soaron these types of projects. Recently, SFTs have beenidentified by the Discovery Channel as one idea for acrossing of very large body of water or maybe evenan ocean. Such a concept has many non-technicalobstacles which may prevent such a project to getfarther than the conceptual stage. However, SFTengineers have identified the major issues to addressto make such a concept work from a technical stand-point (Ostlid, 2006).

494

All of the proposed long sub sea tunnels, suchas the Gibraltar and Bering Straits tunnels, that arenow in the planning stage fall into the category ofbold and visionary tunnels as did the Channel Tun-nel for over a century. ITA has been very active inthe evaluation and promotion of sub sea tunnels,especially the Gibraltar Tunnel, and has summarizedsome of the projects now in the planning stage (ITA2008). The author believes that technology will beeventually developed to make the tunnels currentlyin the planning stage feasible as well.

The Swiss Metro is a bold and visionary con-cept developed for a very high speed transportationnetwork in Switzerland with possible extensions toother parts of Europe. Maglev trains would be pro-pelled at very high speed through tunnels in a partialvacuum. A lesser-known but comparable scheme forNorth America, the American Metro, has been sug-gested for consideration by a railroad dispatcher andrailroad buff, Swartzwelter (2003). A network oftunnels between major cities would be constructedfor very high speed (1000 km/hour) transfer of peo-ple and goods.

These and other very bold, visionary, and dar-ing projects, possibly even the use of undergroundspace on the moon, are part of the forward vision ofour industry. Many may never get past the “What If”idea stage. It is, however, important for our industryto stretch wherever it is appropriate. In a relativesense, this may be similar to previous bold andvisionary thinking by our forefathers which resultedin the Thames Tunnel, the alpine tunnels, and theChannel and Seikan sub sea tunnels.

PLANNING IMPACT FROM ISSUES AND EVENTS OUTSIDE OUR INDUSTRY

There are many issues and events outside of ourindustry that will have a significant impact on ourplanning for tunnels. Some of these issues we do noteven know about yet but one of these is the price andavailability of oil. There have been numerous claimsover the past decades that the world would run out ofoil, or cost so much that other fuels will be neces-sary. This has not happened yet but may happen, ifnothing else, because of greatly increased demandfor oil by developing countries. So there is a muchgreater chance that the world will need substitutetechnology or fuels, possibly to include hydrogenfuel cells or a hybrid. These developments will giveplanners new challenges and new opportunities.

The fuel cell is a concept that was invented in1839 but did not gain any real practical use untilused by NASA for space travel. The concept is beingworked on by many countries including the USA;there are several local and regional agencies whichhave fuel cell or hybrid technology as a test. One ofthe major factors for planners is that when hydrogen

is the fuel, the only emission is water. If vehicleswere to be powered by hydrogen, or a high levelhybrid, adverse emissions are drastically reduced ornearly eliminated. Planners and designers wouldhave to modify their approach to tunnel ventilationduring normal service use while still providingessential fire and life safety services. Such a conceptwould possibly drastically reduce tunnel operationalcosts and possibly even capital cost.

Hybrids are fast gaining popularity both inadditional research and in practical applications byforward thinking and visionary travellers. Some usedifferent fuels which may impact tunnel ventilationsystems in various ways. Some, however chargelarge electric batteries which, when used to propelvehicles through tunnels would drastically reduceventilation demands especially in long tunnels.

It may be preposterous to think that the internalcombustion engine may be replaced or metamor-phosed into something better, but stranger thingshave happened even within the author’s lifetimesuch as the development of television, jet engines,space travel, computers, air conditioning, cellphones, etc. The author is not predicting such atransformation in transportation may take place butour industry should think on what effect it wouldhave on the design of the ventilation system and theoperation of the tunnels if it were to take place.

Another concept that may affect future tunnelplanning is related to propulsion technology such asMagLev or other future propulsion breakthrough wedon’t even know about yet. There are Maglev sys-tems in operation now but not in tunnels althoughthe author understands that a maglev test track inJapan goes through a short section of tunnel. Muchmore development, and reduction in cost, will benecessary to make Maglev, or any other new propul-sion system, worth considering for tunnels. How-ever, there are some promising aspects if the conceptbecomes viable. For instance, the system is capableof being faster and it is environmentally friendly.Moreover, it can negotiate tighter curves and steepergrades both of which may allow tunnels to be shorterand therefore less costly. Whether such a conceptwill be technically feasible or cost effective is yet tobe seen, but consideration of such a concept for boldand daring planning of future tunnels may not beunreasonable.

VISIONARY USE OF UNDERGROUND SPACE

New ideas for improved and greater use of under-ground space are constantly being proposed. Now,thousands of years after the use of caves for refuge,there are very interesting proposals for the use ofunderground space, especially from Japan. Theseinclude multi-modal complexes with offices, living

495

quarters, public meeting areas, sports arenas, recre-ation, schools, research facilities, etc. all connectedto the rest of the city at multi-modal stations.

There are abundant locations where products(such oil, gas and even wine) are already beingstored in bulk storage caverns. There are wine cel-lars created for storage and entertainment. In KansasCity in the United States, there are existing minedopenings with over 450,000 square meters of spacethat is leased to various users for storage, officespace, etc. New technology, concepts, and designs toprovide as much sunlight to underground facilitiesas possible are being developed (Bennett 2008).Major underground space facilities require creativevision but can be planned boldly just like any otherunderground project.

Beneficial use of underground space is not justfor the developed countries. Hand-dug undergroundcellars and bins have been used for centuries tosafely store food and other goods. This may still beone of the best and least expensive ways to improvethe quality of life in the developing countries.Underground spaces, properly designed, can be builtwith local labor and limited outside resources to pro-vide good and safe environment for storage of water,food and other products and equipment. If designedand constructed properly they are earthquake resis-tant and energy efficient. As part of the work ITA isdoing as an NGO with the United Nations, ITA hasproposed that the development of mined or earthsheltered construction and underground space beincluded as part of the effort to create sustainablerural communities, especially in areas with extremeclimates.

CONCLUSIONS

As the growth of cities, particularly the megacities,continues, they will all need abundant infrastructureand the demand for the Underground will be enor-mous. In fact, even if a fraction of the needed infra-structure is funded, the capacity of our industry willbe challenged. We would be challenged to build therequired infrastructure fast enough. There simplymay not be enough planners, designers, and contrac-tors to do the work.

Tunnels and underground space play a domi-nant role in our standard of living and in the preser-vation of the environment. Moreover, they areincreasingly recognized as being environmentallyand socially friendly and energy efficient. Under-ground Space is Good for Sustainability and is anAlly in the fight against Global Warming.

In past years, planners did not have the luxuryof the many tools and techniques that exist today.We should be very proud of our forefathers who hadgreat vision and whose ideas were not achieved in

their lifetime. Technology development was so slowfrom about 1850 to 1950 that their ideas did notmaterialize until technology made their schemesfeasible. Now (2008), technology is keeping up withour ideas and vision which can be implemented rela-tively quickly. Accordingly, planners should beaggressive and bold in their plans for tunnels andunderground space.

During the lifetime of the project you are plan-ning today, new technology and/or environmentallydriven concepts will develop that may significantlyaffect your project.

The principles discussed for traditional tunnelsalso apply to the more noble use of undergroundspace for living, working, and storage. Nor are theyconfined to application only to developed countries.ITA has recommended that the development of earthsheltered construction and underground space beincluded as part of the effort to create sustainablerural communities, especially in areas with extremeclimates.

Owners and planners should use risk manage-ment principles from the very first time a tunnelsolution is considered and carry out systematic riskmanagement evaluations throughout planning,design and construction. These same principlesshould be used to systematically develop and imple-ment value engineering and new opportunities, espe-cially those thinking out of the box. This isparticularly true for long tunnels that have abundantuncertainties. Our industry can not afford to wait fordecision makers to become aware of the sustainabledevelopment benefits of the Underground. Insteadwe must be pro-active and let the world know thatsustainability is not possible without infrastructureand that, often, the best form of infrastructureinvolves the underground.

Owners should develop ways to account for afinancial credit resulting from environmental advan-tages especially for long transportation tunnels thataccrue enormous environmental benefits to society.These environmental cost advantages should beincorporated into cost ranges that take into accountthe long service life of the tunnels by making thedecision on Life Cycle Cost and Benefit concepts,not initial capital cost. It should always be remem-bered that tunnels are an investment, not a cost.

There are numerous bold and visionary projectsalready completed or under consideration. There areeven daring projects that may not get built but whichstretch our imagination of what is possible withunderground space. The tunnel and undergroundindustry is very creative and our ability to innovatehas been proven many times. Owners and plannersshould have faith in the tunnel and undergroundindustry. The industry will be up to any challenge soplanners can plan boldly with vision.

496

REFERENCES

Bennett, David J., 2008. “Beyond Sustainability:The Case for Environmental Architecture,” http://djbarchitects.home.att.net/sustainablebase.htm,16 pp, Accessed January 14, 2008.

ITA, 2007. Underground Space, an UnexpectedSolution to Promote Sustainable Development.Joint UN-ITA Workshop conducted at the UnitedNations, December 14, 2007, In Press at ITAWebSite: www.ita-aites.org.

ITA, 2008. Challenging Tunnel Projects Worldwide,ITA WebSite: www.ita-aites.org, Accessed Janu-ary 14, 2008.

Ostlid, Havard, 2006. Tunnels Between Continents?Possible or Not Possible? Proceedings of ITATraining Workshop, Seoul, Korea, ITA WebSite:www-ita-aites.org.

Parker, H. W., 1976. Field-Oriented Investigation ofConventional and Experimental Shotcrete forTunnels, Thesis submitted to the University ofIllinois at Urbana-Champaign, Illinois, in 1976 inpartial fulfillment of the requirements for thedegree of Doctor of Philosophy in Civil Engineer-ing, 628 p.

Parker, Harvey W., 1999. The Development of Hardand Soft Ground Tunnels, Proceedings, TenthAustralian Tunnelling Conference, KeynoteAddresses and Asia-Pacific Forum Volume, TheAustralasian Institute of Mining and MetallurgyPublication Series 1a/99, Victoria, Australia,1999, pages 25–31.

Parker, Harvey W., 2001. Early Developments ofShotcrete and Steel Fiber Shotcrete, Proceedings,Eighth Engineering Foundation Conference onShotcrete, Campo do Jordo, Brazil, 1999, UnitedEngineering Foundation Press, ASCE, 2001,pp 240–258.

Parker, H.W., 2004. Underground Space: Good forSustainable Development, and Vice Versa,” Pro-ceedings, ITA-AITES World Tunnel Congress,Singapore, May 2004.

Parker, H.W. and Semple, R.M. 1972. Developmentand Large-Scale Testing of Innovative TunnelSupport Systems, Proceedings, North AmericanRapid Excavation and Tunneling Conference,American Institute of Mining, Metallurgical andPetroleum Engineers, Chicago, Illinois, 1972,p. 469–497.

Parker, Harvey W. and Reilly, John, 2008. Risk andLife Cycle Costs and Benefits for UndergroundProjects, Proceedings, North American TunnelingConference, San Francisco, SME, Denver(In Press).

Reilly, J.J. 2006. Cost Estimating, Probable Cost,Risk Identification and Risk Management forInfrastructure Projects. Proceedings AmericanPublic Transportation (APTA) Conference,June 2006.

Reilly, J.J. and Parker, H.W. 2007. Benefits and Life-Cycle Costs of Underground Projects. Proceed-ings, AITES-ITA World Tunnel Congress, Pra-gue..Vol. 1: 679–684.

Stack, B., 1982. Handbook of Mining and Tunnel-ling Machinery, John Wiley and Sons: Chichester,742 pp.

Swartzwelter, Brad, 2003. Faster than Jets, AlderPress, Kingston, WA, USA, 188 pp.

Wallis, Shani, 2007. Personal Communication.

576

Probable Cost Estimating and Risk Management

John ReillyJohn Reilly Associates International, Framingham, Massachusetts

ABSTRACT: Costs of complex/underground projects have been consistently underestimated and deliveryon time and on budget has been problematic. This has been attributed, to inadequate cost estimates and uniden-tified and/or insufficiently managed risk. Use of probabilistic cost estimates like the Washington State Depart-ment of Transportation’s CEVP® process (developed by the author and others) has produced more realisticcost estimates, explicitly identified and quantified risk (as input to risk management plans) and early aware-ness of key project issues. This paper updates previous work (NAT 2004), adds implementation detail forprobalistic cost estimating as used by WSDOT, FHWA and other agencies and will address key issues in riskmanagement, as related to probalistic cost estimates.

INTRODUCTION

Good project management requires, among otherkey elements, good cost and schedule estimates, aswell as a process to identify and deal with uncer-tainty and management of risk. These elements arethe focus of this paper.

THE COST ESTIMATION PROBLEM

Even when a project is well planned and managed,in a significant number of cases project conditionschange and problems arise. Of these, politicalchanges seem to have the most significant effect(Salvucci 2003). These changes and problems haveresulted in significant undesirable consequenceswhich include cost and schedule over-runs, resourcecompetition between projects, negative media atten-tion and, consequently, public mistrust.

Thus we find that the public is skeptical of ourability, as a profession, to accurately estimate thefinal costs of large, complex public projects and isalso skeptical of our ability to manage these projectsto established budgets. Questions they have askedinclude:

• “Why do costs seem to always go up?”• “Why can’t the public be told exactly what a

project will cost?” and,• “Why can’t projects be delivered at the cost

you told us in the beginning?”

Our inability to answer these questions consis-tently is a consequence of many factors—includinginadequate cost estimating procedures and our priorinability to correct these poor estimating practices(Flyvbjerg et al. 2002). Additionally, the effects ofpoor project management and poor communicationwith the public has further added to the problem—

resulting in unfortunate results, including negativevotes for proposed transportation funding.

Many government agencies, including the Fed-eral Highway Administration (FHWA) and someState Departments of Transportation have recog-nized this problem and in response are now requir-ing risk-based cost and schedule estimating, as wellas formal risk management plans (FTA, 2003; FTA2004, FHWA 2006).

Examples of Poor Cost Information and Estimation

One international survey (Reilly and Thompson,2000) found that specific, relevant cost informationwas usually unobtainable. Little objective historycould be found, including findings that would sup-port recommendations for improvement. Because ofthe difficulty in obtaining hard data, firm conclu-sions could not be reliably drawn but the followingfindings were reported by the owners:

• There are significant cost and schedule over-runs suggestive of poor management in at least30%, and possibly more than 50%, of theprojects.

• It appears that the factors that most directlyinfluence success or failure are (a) expertiseand policies of the owner and (b) procurementprocedures.

• The professional teams engaged in projectswere judged, by the owners, to be competentleading to the conclusion that problems inpoorly performing projects may lie primarilywith the ability of the owner to effectively leadand manage the project process.

• Risk mitigation was not well-understood orapplied, even in elemental ways. This was con-sidered to be a promising area for development,

577

in particular as it related to cost over-runs andunforeseen events.

• Reported cost and performance data—espe-cially for “good” results—should be treatedvery cautiously.

• Consistent, complete and relevant data are veryhard to get and almost impossible to validateafter project completion.

• Conclusions based on reported cost data, unlessthe conclusions are grossly evident (e.g., meta-findings), should also be treated with caution.

Other studies confirm the problem (Flyvbjerg et al.2002). Flyvbjerg surveyed 258 projects spanning70 years and found that the problem of accurate costforecasts has been chronic for that time period. Keyfindings were:

• 9 out of 10 transportation projects underesti-mated costs with an average overrun of +28%

• Road projects averaged +20% higher thanestimated

• Tunnel and Bridge averaged +34% higher• Rail projects averaged +45% higher than

estimated

Specific examples include several of the FTA Dem-onstration projects such as Tren Urbano in PuertoRico—$1,285 million over initial budget (+133%),the Silver Line Transitway Project in Boston—$286million over budget (+90%), the London JubileeLine Transit Project—2 years late and £1.4 billionover budget (+67%), the Channel Tunnel RailProject—£3.7 billion (+80%) over budget, Den-mark’s Great Belt Link rail and road link (+54%),the 2003 Woodrow Wilson bridge tender in Virginia72% over estimate and, Boston’s Central ArteryProject—many billions over the initially publishedcost numbers (which were extremely unrealistic)and years late.

Note: the cost percentage number in parenthe-ses in the above paragraph indicates the final cost ofthe project divided by the budget which was com-municated at time of decision to proceed. Thisimportant because it is the “number” that the publictends to remember and the media reports. However,it does not include new scope and other changes,which might be quite legitimate. It does include poorinitial cost estimates and/or poor estimation of riskand other factors—including poor management, theeffects of external events and political changes ortransitions (Salvucci, 2003).

Of major concern is that, as an industry, wehave not corrected the “chronic cost underestima-tion” of such projects—if we had done so, therewould have been a uniform number of results

equally over budget as under budget. This problemhas existed for over 70 years, as shown by Flyvbjerg,whose conclusion is that the problem is both aninability to estimate accurately, a bias to estimate onthe optimistic side and political misrepresentation.

The Influence of Variables

The final, definitive cost and schedule of a projectcannot be predicted accurately because the projectcan, and will, be affected by a number of variables.These variables include nature (e.g., ground condi-tions, weather), technology (e.g., design, methods,equipment, materials) and human (e.g., labor, pub-lic, politics, regulatory agencies, funding/insurance/bonding agencies, market conditions). These vari-ables cause uncertainty in cost and schedule throughvariations in those project conditions assumed forthe estimates (e.g., average unit rates, progress rates,and escalation rates) and through uncertain impactsfrom unplanned events (deviations from thoseassumptions).

Early Information and Uncertainties

Early in project development when project estimatesare initially developed, information on these vari-ables is typically limited, but a process for identify-ing and managing potential problems should bedeveloped at this time, when the problems are easierto resolve, if identified.

If these uncertainties are not explicitly includedin the estimating process, and if inevitable estimat-ing biases are not corrected, project cost and sched-ule estimates are likely to be inaccurate, consensuswill be difficult to achieve, and it will not be possibleto answer critical questions such as: “Which projectalternative is best?”; “What scope is actually afford-able or will actually be built?”; “Is the current esti-mate high or low relative to what the actual cost andschedule will ultimately turn out to be, and by howmuch?”; “What should the funding/budget and mile-stone dates be for this project?”; “How can theproject cost and schedule best be controlled?

POTENTIAL SOLUTIONS

A variety of approaches have been developed toattempt to provide better cost and schedule esti-mates. These various approaches generally differ inthe following areas:

• Traditional contingency-based deterministic(single value) approaches vs. probabilisticapproaches—either combined uncertainty oritemized uncertainty, e.g., individual risks;

• Separate and unlinked cost and schedule mod-els vs. integrated cost and schedule models;

• Varying levels of detail and approximation; and

578

• Input assessment methods (e.g., cost data base vs.project-specific judgment of technical experts).

The author, clients and colleagues (WSDOT 2007,Roberds and McGrath, 2005; Grasso, 2002;) believethat a flexible (depth and breadth of detail anddegree of approximation), probabilistic, risk-basedapproach using an integrated cost and schedulemodel is the most appropriate way to quantifyuncertainties for complex projects and to guide riskmanagement in order to better define and controlcosts and schedules.

Example of One Approach

Such an approach forms the basis of the Washing-ton S ta te Depar tmen t o f Transpor ta t ion’s(WSDOT) Cost Estimate Validation Process(WSDOT, 2007) which is used on all WSDOTprojects over $25 million has been used on a num-ber of Federal Highway Administration and StateDepartment of Transportation programs and otherAgency projects (e.g., Alaska Railroad, TorontoWaterfront Revitalization). The approach has alsobeen used to quantify uncertainty in programmaticmeasures such as program expenditure and cashflow and for programs consisting of a large numberof individual projects, each of which are uncertainbut often related to some degree. Specifics of thedevelopment of this approach follow.

WSDOT’S COST ESTIMATE VALIDATION PROCESS—CEVP

In January 2002, the Washington State Secretary ofTransportation was questioned by a State Senatorabout the poor reliability and history of increases ofcost estimates for a large project. WSDOT managersand key consultants—“the core team” (Reilly et al.,2002)—were asked to develop a better cost estima-tion process. As part of defining “the problem,” areview of relevant data led to the following findings:

1. There is a general failure to adequately recog-nize that an estimate of a future cost or sched-ule involves substantial uncertainty (risk),

2. Uncertainty must be included in cost estimating,3. Cost estimates, must be validated by qualified

professionals, including experienced construc-tion personnel who understand “real-world” bid-ding and construction,

4. Large projects often experience large scope andschedule “changes” which affect the final cost.Provision for this must be made in the cost esti-mates and management must deal competentlywith these changes.

WSDOT decided to act on these findings by:

1. Developing an improved cost estimatingmethodology,

2. Incorporating cost validation, risk identificationand management,

3. Openly and reliably communicating “ranges ofprobable cost” to public, media and politicaldecision makers.

WSDOT’s strategy was to deal openly with the pro-cess of public infrastructure cost estimating so thatthe public would better understand, and be betterinformed, as project manages and elected officialsmake critical project funding decisions. The chal-lenge was to develop a valid procedure to do this.WSDOT decided to open the “black box” of estimat-ing and present a candid assessment of the range ofpotential project costs, including acknowledgmentof the uncertainty of eventual project scope, theinevitable consequence of cost escalation due toinflation, and other major risks.

The WSDOT Process

The WSDOT team developed a specific manage-ment-cost-risk assessment tool which was called the“Cost Estimate Validation Process” or CEVP® . Thedraft procedure included:

1. A cost validation process adapted from theBoston Metrowest Water Supply Tunnelproject,

2. Inclusion of the impacts of risk and opportunityevents derived from procedures previouslydeveloped for infrastructure tunnel projects (Ein-stein 1974; Anderson, Reilly and Isaksson 1999;Grasso et al. 2002),

3. Use of independent subject matter experts, par-ticularly those who have been responsible forconstruction of these projects in a hard-moneybid environment.

The basic approach requires a peer-level review, or“due diligence” analysis, of the scope, schedule andcost estimate for a project and then incorporation ofuncertainty (uncertainty includes both risk andopportunity) to produce ranges of probable cost andschedule. Specific objectives of the method are toevaluate the quality and completeness of the “basecosts” together with the inclusion of inherent uncer-tainty (risk and opportunity) in the estimate. Riskmitigation can also be included.

WSDOT launched the CEVP® program with amajor commitment of its personnel, including func-tional and project staff, staff from project partners,members of the consultant teams already working on

579

some of the larger projects and, the core CEVP devel-opment team. To this were added external specializedconsultants including a group of very senior engineer-ing, construction and cost estimating specialistsdrawn from around the country (see “Acknowledg-ments” at the end of this paper). Figure 1 shows basecost and the range of probable cost histogram.

Key Concepts

As the CEVP methodology emerged, several keyprinciples were identified. Among these were:

1. Avoid single number estimates. Recognize thatat any point in the development of a project,from initial conceptualization through the end ofconstruction, an estimate will require selecting arepresentative value to characterize many factorsthat are inherently variable. These variable fac-tors will include issues that have been identifiedand quantified (the known/knowns), those thathave been identified but not yet quantified (theknown/unknowns) and those that have not yetbeen identified (the unknown/unknowns). Somefactors will be controllable by design or by theowner, some will not. But all of these contribut-ing factors are fundamentally uncertain and needto be treated as such.

2. Use a collaborative assessment environment thatcombines high levels of critical external peerreview expertise, particularly in construction andestimating construction in a bidding environ-ment, with appropriate roles and responsibilitiesfor the Project Team. Project Teams and ownersare (and should be expected to be) biased. Theyare generally too optimistic about the project andwant to see it advanced, funded and built. Bal-ance this bias with independent subject matterexperts, peers and others with valued experiencethat is based on experiences separate from thespecific project.

3. Acknowledge that both cost uncertainty andschedule uncertainty are major contributors toproblems with project estimating, and incorpo-rate both in the evaluation methodology.WSDOT foresaw the clear advantage, in fact thenecessity, to integrate the effects of cost andschedule uncertainty. CEVP® was developed:– To incorporate quantified uncertainty for

both risk and opportunity factors,– To identify these factors using an aggre-

gated-component approach that separatedcomponents whose cost and/or durationcould be considered separately and,

– To integrate cost and schedule usingappropriate analytical methods.

4. Be practical and use common sense notions ofrisk descriptions and quantification. The CEVP®

method was to be completely rigorous and treatuncertainty in ways that acknowledged correla-tion, independence and other probability princi-ples. However, the sources of information anddefinition of uncertainty were likely to encom-pass a range which might extend from highlyquantified issues to those where subjective opin-ion from the contributors was all that would beavailable. This range of uncertainty data neededto be captured objectively.

5. Produce project output that could be under-stood by the ultimate audience, the public. Thisled, ultimately, to a focused approach that pre-sented the concepts of cost ranges, probability,risk and opportunity and risk management tothe public, media and political decision makersin June of 2002. The public, media and politicaldecision makers accepted these concepts sur-prisingly quickly and without major comments.

CEVP®—PROCESS AND RESULTS

CEVP® develops a probabilistic cost and schedulemodel to define the probable ranges of cost and

Figure 1. Future costs are a “range of probable cost”

580

schedule required to complete each project. Theresults of the assessment are expressed as a probabledistribution of cost and schedule values for theproject as shown in Figures 1 and 2. The CEVP pro-cess (Reilly and Brown, 2004):

1. Critically examines the Project estimate to vali-date all cost and quantity components usingindependent external professionals.

2. Removes all “contingency” and allowances forunknowns, then

3. Replaces the contingency and other approximat-ing allowances with individually identified andexplicitly quantified uncertainty events (risksand opportunities).

4. Builds a model of the project—normally in anExcel spreadsheet using a flow-chart of keyplanning, design, permitting and constructionactivities. Included are quantification of cost andcritical path schedules. The model assigns thequantified uncertainty events to activities withthe associated probabilities and impacts.

5. Runs a simulation to produce the projected“range of probable cost and schedule” andreports the results.

CEVP® Workshop Elements

Key elements of the CEVP® workshop are

1. Develop a model of significant project elementsreflected in a “flow chart,”

2. Cost validation—determination of “base cost”and,

3. Risk elicitation—defining the probability andimpacts of risk events.

The fourth element is the subsequent communicationof results to the public, media and political decision-makers. The basic CEVP® process has been describedpreviously in several papers (Reilly and Brown 2004,Reilly 2005) so the following text will describe theflow chart, cost validation and risk elicitation.

The Project Flow Chart

The Project Team provides a detailed description ofthe expected project plan, with the major activitiesand their associated costs and durations. From thisinformation, the team develops a project flow chartthat represents the sequence of major activities to beperformed in the project. Major decision points(e.g., funding decisions) and project milestones, asdescribed by the Project Team, are explicitly repre-sented in the flow chart. The base costs and dura-tions, as well as any related major uncertainties orcorrelations for each activity are entered on the flowchart using values as confirmed or defined by thebase cost review team.

Cost Validation

The cost part of the workshop is led by a managerwith program delivery experience, supplemented byteam members with both design and real-world con-struction experience. The use of personnel withexperience in contractor’s methods is necessary tobring that perspective into the cost review for a well-shaped determination of “base cost”—the cost if “allgoes as planned and assumed” without contingency.The process consists of the following:

1. The project team first briefs the CEVP® spe-cialists on the detailed scope of the project andidentifies cost and schedule risks that that havebeen included in the project estimate.

2. The CEVP® cost specialists discuss the cost esti-mate with the project team, reviewing what theestimate represents and the basis of its develop-ment. They discuss what metric has been used tocalibrate the estimate and what contingencieshave been included in the estimate.

3. A review of the scope of the project is completedwith the project team on an element by elementbasis to assure that all elements and phases ofthe project have been accounted for.

4. The estimate is reviewed to assure that itemssuch as: right of way, mobilization, permitting,

Figure 2. Model results—probability vs. probable cost

581

mitigation, temporary facilities and utilities, con-struction phasing requirements, seasonal con-straints, cuts/fills, hazardous material issues,archaeological issues, storage and disposal ofmaterial, haul distances, compaction and testing,protection of work, testing of mechanical andelectrical systems, occupancy permits, demobili-za t ion , e tc . have been recognized andaddressed—from a cost standpoint.

5. The schedule for the project is also reviewed—Is it realistic? Does it consider adequate timefor mobilization? Set-up of temporary facilitiesand utilities? Construction permitting? Con-struction phasing? Dealing with differing siteconditions? Traffic or operational issues? Sea-sonal constraints? Site access limitations? Test-ing of piping? Electrical and signals? SCADAsystems?, etc.

6. Unit prices and production rates that have beenassumed for the major items of work arereviewed, asking if the production numbers (thebasis of the units costs) are reasonable and ifthere are any risks that those unit prices may nothave taken into account—such as high groundwater or the presence of organic material.

7. The contingency that is included in each unitprice—or the entire estimate—is identified andremoved from the cost estimate. This is done inorder to develop the “base cost” of the project(the contingency is subsequently replaced by theprobable cost of risk and opportunity events).

8. During the discussions, and upon completion ofthe above review, items of work that may bemissing, over- or under-estimated are identifiedand recorded. Estimates for missing items aredeveloped and recommendations for adjust-ments are made. Finally, an agreed “base cost” isdetermined. This becomes the base to which thecost of potential risk and opportunity events areadded by the cost/schedule uncertainty model.

Risk Identification and Quantification

The risk identification and quantification is led by anexperienced risk elicitator/analyst who is familiarwith uncertainty theory, de-biasing techniques andthe structure of a subsequent cost and risk model.Other workshop participants include representativesfrom the project team who have familiarity with theplans, strategies, assumptions and constraints on theproject, plus the Subject Matter Experts (SME’s)who bring an independent perspective on importantareas of project uncertainty.

The identification and quantification of uncer-tainties requires a balance of project knowledge, riskanalysis expertise, cost estimating expertise, andobjectivity. Project knowledge and the independentexpertise of SME’s are essential to identify the

uncertainties. Risk analysis expertise is required tocapture balanced information on risk and modeluncertainties. The goal of the risk workshop is toidentify, quantify and model the uncertainty inproject cost and schedule. The risk identificationprocess—preparation and workshop—includes thefollowing activities:

1. Introduction to the participants of principles ofuncertainty (risk and opportunity) assessment

2. A preliminary list of risks and opportunities,generated by the project team

3. Workshop identification of potential risks andopportunities. This is done in an open brain-storming process that typically begins with aprior list of potential uncertainties from theproject team, lists from similar projects andother sources. In the workshop, it is necessary toprovide a critical environment that allows forthis initial information to be combined withother suggested risks. As a practical matter, theteam should identify a screening criteria to helpproduce a prioritized list of significant cost andschedule risks.

4. Characterization of potential risks and opportu-nities. This process combines subjective andobjective information to identify the conse-quences to the project if each of the risks were tooccur. Typically there are varying opinions onthe range of consequences, such as increasedcost or schedule delay. The risk elicitator isresponsible for guiding the group to an appropri-ate agreement regarding the consequences of therisks—i.e., probability of occurrence andimpact. Independence and correlation amongrisks is also defined, positively or negatively orconditionally.

The risk and opportunity events that are the outputof the workshop should be defined to be independentas far as possible. When this is not possible, thedependencies among events must be defined andaccounted for. In addition, each risk or opportunityevent must be allocated to the project activities thatare affected by it or, if a given event affects multipleproject activities, significant correlations amongoccurrences need to be addressed. Significant uncer-tainties and correlations among event impacts alsoneed to be defined. This information is incorporatedin the cost and schedule to produce the modelresults.

Risk elicitation in the workshop is an iterativeprocess that combines subjective and objectiveinformation. Uncertainty characterizations and prob-abilities are defined simultaneously to provide rea-sonable, practical descriptions of uncertainty.

582

Completion of the CEVP® process

The “range of probable cost and schedule” is deter-mined by combining base costs determined from thecost validation with the risks identified in the riskworkshop in a Monte-Carlo model which:

1. Integrates base costs with risks and opportuni-ties (uncertainties) in a probalistic model

2. Reports the results as a “range of probable costand schedule”

Results of the model analysis are presented as costand schedule probability distributions, usually in agraphical form (Figures 2, 3) with supporting tabula-tions of characteristic statistics. First order descrip-tions and models are sufficiently accurate and areused for most projects. These distributions candescribe a variety of situations of interest including:

1. Current dollar (time of assessment) cost andyear of expenditure cost ($YOE)

2. Fully funded or partially funded scenarios3. Comparative design options4. Probable date of completion for the project5. Probable schedule to meet project milestones

The specific form of the reported results canvary depending upon need and the results can beused for a number of applications, including:

1. Project assessment/management re-direction2. Risk management plans3. Data input to value engineering workshops4. Integrated management of multiple projects5. Better internal and external communications6. Financial management options and alternatives

CEVP® is iterative in nature and represents a“snapshot in time” for that project and under theconditions know at that point. Changes to optimizecost or reduce risk can be verified in a reassessmentof the project and model update. An example for acomplex project in Seattle comparing 2003 results to2002 shows a reduction in both probable cost andrange of uncertainty by management of scope andrisk (Figure 3).

RISK MANAGEMENT

Early, strategic risk management is one of the mostimportant tools for managing cost and schedule.Referring to London’s Jubilee Line Transit Projectwith a cost overrun of +67% the Secretary of State’s

Figure 3. Improvement in probable cost for successive CEVP® workshops

Figure 4. General risk management approach after CEVP®

583

Agent (oversight consultant) stated: “Time and costoverruns could have been minimized with a moreestablished strategy at the very beginning of theproject” (Arup, 2000).

But, how to determine the “more establishedstrategy at the very beginning of the project”?

A key output from the CEVP® assessment is aranked listing of the risk and opportunity events con-tributing to the uncertainty in a particular estimate.The ranked risk table presents the most importantrisk issues, along with a measure of their contribu-tion to the total uncertainty in the estimate. The vari-ety of risks, including technical risks, policy risks,environmental risks, construction risks, etc. can betreated in a consistent way using this data.

One of the not-so-incidental benefits of CEVP®

is that it provides an explicit quantification of poten-tial risk and opportunity events that could impact theproject’s cost and schedule. From this quantified riskprofile, risk management plans can be developedearlier in the project life-cycle. Risk managementprocedures are well understood and many referencesare available (Einstein et al., 1974; Roberds andMcGrath 2005; Grasso et al. 2002; Isaksson, 2002).

CURRENT DEVELOPMENTS

CEVP® is proving to be a useful process for estimat-ing and communicating ranges of probable costs andschedules, as well as explicitly identifying and quan-tifying risks for large, complex projects early in theplanning and design phases. This produces betterinformation that the public and elected officials andcan be used to make more informed decisions, whileallowing engineers to better manage these projects.

WSDOT has internalized the CEVP® processand uses it, and a simpler, cheaper “Cost RiskAssessment” process, for many of its projects. TheU.S. Federal Transit Administration (FTA) and theFederal Highway Administration (FHWA) have eachinvestigated CEVP and similar processes and haverun demonstration projects. They have concludedthat a probalistic cost-risk process, such as CEVP®

or an alternative, should be used for most large,complex transportation projects.

As of this writing further demonstrationprojects and educational seminars are underway andseveral government Agencies are beginning torequire the process in their upcoming projects (e.g.,Florida, Toronto Waterfront) and internationalenquiries following presentations of the CEVP pro-cess have shown interest in its application.

FINDINGS

1. WSDOT recognized the value that cost validationand risk assessment yields in the determination of

the “range of probable cost” including explicitpotential risk events.

2. WSDOT found that the CEVP® results alloweda more intuitive communication with the publicwhich better related to “…what people alreadyknow.”

3. WSDOT found that CEVP® focuses early atten-tion on the significant cost and schedule risks fora project and increases the project team’s aware-ness of risk.

4. Use of experienced external subject matterexperts, who have constructed similar projects,is good value and gives an independent check onkey assumptions.

5. Because risks are explicitly defined, a risk man-agement plan can be quantified earlier. Thisallows significant management and control ofcost and schedule earlier in a project and allowsa more explicit communication of cost andschedule (and changes thereto) with the publicand key political decision makers.

6. WSDOT recognized that CEVP® is not a“magic bullet” or a “quick fix.” WSDOT there-fore committed to improve its cost estimatingprocess by implementing the CEVP® programon a state-wide, long-term basis and trainingstaff in the technique.

7. Presentation of CEVP® to the industry, includ-ing FHWA representatives, resulted in a review,assessment and recommended use of probalisticcost estimating processes by that and relatedagencies.

RECOMMENDATIONS

1. Owners of complex infrastructure and under-ground projects should consider using a probal-istic cost–risk type process for cost estimatesand risk identification.

2. Periodic updates to the model should be used toassist with, and explain rationally, projectchanges. They can, additionally, evaluate andcompare alternatives.

3. The process should be used to assist owners indetermining a more realistic “range of probablecost” which can enable better communicationabout, and management of, these projects.

4. This is only possible if the owner truly wants toknow the realistic “range of probable costs,” isprepared to communicate this to the public anddecision makers and then will manage theprojects to the subsequently established budgets.

5. It is recognized that significant concerns havebeen raised that, if we tell the public the morerealistic probable costs, which tend to be greaterthan other estimates, those projects may not befunded and authorized. Such concerns imply that

584

we cannot trust the public with the real costinformation. If true, such a position has moralimplications that the profession needs to address.

6. A risk management plan should be developedand implemented using the explicit, quantifieddefinition of potential risk events.

ACKNOWLEDGMENTS

This paper follows previous papers and presenta-tions by the author and colleagues describing theCEVP® process. References can be found on theWSDOT website at http://www.wsdot.wa.gov/projects/projectmgmt/riskassessment.

Material from Michael McBride (formerly withMWRA Boston); Dwight Sangrey, Bill Roberds andTravis McGrath of Golder Associates (see refer-ences following) was used in this paper. The authoris grateful for this input.

REFERENCES

Arup, 2000, “The Jubilee Line Extension, End-of-Commission Report by the Secretary of State’sAgent” UK Department of the Environment,Transport and Regions, July.

Einstein, H.H. and Vick, S.G. 1974, ‘Geologicalmodel for a tunnel cost model’ Proc Rapid Exca-vation and Tunneling Conf, 2nd, II:1701–1720.

Flyvbjerg, B., Holm, M.S., and Buhl, S. 2002,“Underestimating Costs in Public Works Projects:Error or Lie?” Journal of the American PlanningAssociation, Summer, 2002. Vol. 68, Issue 3;pg. 279–296.

FHWA, 2006, “Guide to Risk Assessment and Allo-cation for Highway Construction Management,”Report FHWA-PL-06-032.

FTA, 2003, Operating Guidance Memorandum No.22, “Risk Assessment Procedures, Requirementsand Report Formats for PMO Contractor Deliver-ables and Services,” US DOT, November 10th.

FTA, 2004, “Risk Assessment Methodologies andProcedures,” Project No. DC-03-5649, WorkOrder No. 6, May.

Grasso, P., Mahtab, M., Kalamaras, G. and Einstein,H. 2002, ‘On the Development of a Risk Manage-ment Plan for Tunnelling,’ Proc. AITES-ITADownunder 2002, World Tunnel Congress, Syd-ney, March.

Isaksson, T., 2002, ‘Model for estimation of timeand cost, based on risk evaluation applied to tun-nel projects,’ Doctoral Thesis, Division of Soiland Rock Mechanics, Royal Institute of Technol-ogy, Stockholm.

Lockhart, C. and W. Roberds, 1996, “Worth theRisk?” in ASCE Civil Engineering, April 1996.

Reilly, J.J., Isaksson, T. and Anderson, J., 1998,‘Tunnel Project Procurement and Processes—Fundamental Issues of Concern’ Conference on“Reducing Risk in Tunnel Design and Construc-tion,” Basel, December.

Reilly, J.J. and Thompson, 2001, ‘International sur-vey, 1400 projects,’ confidential report.

Reilly, J.J., 2001, ‘Managing the Costs of Complex,Underground and Infrastructure Projects,’ Ameri-can Underground Construction Conference,Regional Conference, Seattle, March.

Reilly, J.J., 2002, w. McBride, M., Dye, D. andMansfield, C.—Guideline Procedure. ‘Cost Esti-mate Validation Process (CEVP)’ WashingtonState Department of Transportation, January.

Reilly, J.J., 2004, w. Brown, J. “Management andControl of Cost and Risk for Tunneling and Infra-structure Projects” Proceedings, World TunnelingConference, International Tunneling Association,Singapore, May.

Reilly, J.J., McBride, M., Sangrey, D., MacDonald,D. and Brown, J., 2004, “The development ofCEVP—WSDOT’s Cost-Risk Estimating Pro-cess” Proceedings, Boston Society of Civil Engi-neers, Fall/Winter ed.

Reilly, J.J., 2005, “Cost Estimating and Risk Man-agement for Underground Projects,” Proc. Inter-national Tunneling Conference in Istanbul, May,Balkema.

Reilly, J.J., 2006, “Risk Identification, Risk Mitiga-tion and Cost Estimation,” Tunnelling andTrenchless Construction, April 2006.

Roberds, W. and McGrath, T., 2005, “QuantitativeCost and Schedule Risk Assessment and RiskManagement for Large Infrastructure Projects,”Project Management Institute October.

Salvucci, F.P., 2003, “The ‘Big Dig’ of Boston, Mas-sachusetts: Lessons to Learn,” Tunnels and Tun-ne l l ing , Nor th Amer ica , May and Proc .International Tunnelling Association Conference,Amsterdam, April.

WSDOT, 2007, Guidelines for CEVP® and CostRisk Assessment, Oct 31 2005, updated 2007http://www.wsdot.wa.gov/projects/projectmgmt/riskassessment.

large diameter soft ground bored tunnel review

Documents